Power distribution network fault recovery method and system based on load side island division

Abstract

The application relates to a power distribution network fault recovery method and system based on load side island division, belonging to the field of power systems, which comprises the following steps: obtaining a connected graph based on a power distribution network topology structure and performing edge weight setting, searching for an initial island from a first-level load, taking the initial island as a search starting point to search for remaining loads into the island by using edge weights, obtaining a power distribution network island division result and a target function thereof; calculating power supply redundancy of each island in the island division result, performing island fusion based on the power supply redundancy of each island to obtain an island fusion result; obtaining an island fusion result target function based on the island division result target function; when the power distribution network fails, obtaining a maximum load recovery amount of the power distribution network based on the island fusion result and a constraint condition to recover power supply to the load. The islands are divided, the power supply redundancy of each island is considered, the connection between the islands is established, the connectivity of power supply to important loads is ensured, the power supply to important loads is preferentially recovered, and the coupling between the islands is increased.

Description

Technical Field

[0001] This invention belongs to the field of power system technology, specifically relating to a method and system for distribution network fault recovery based on load-side islanding. Background Technology

[0002] In recent years, extreme weather events have become increasingly frequent. To enhance cities' ability to cope with extreme disasters, improving the resilience of urban power distribution networks and ensuring power supply to critical loads has become increasingly important. As the end point of the urban power grid, the power distribution network directly serves various electricity users and is a crucial link in power supply. Utilizing local distributed power sources to quickly restore power to critical users within the distribution network is an effective means of reducing power outage losses and improving the network's resilience. Power outages to critical loads can cause personal injury and death, resulting in significant political and economic losses. Ensuring the normal and continuous power supply to critical loads after a fault is of great significance for protecting people's lives and production, mitigating disasters, and promoting social development. Ensuring the continuity of the path from the source to critical loads is a key technology for improving the distribution network's disaster response capabilities.

[0003] As distributed generation (DG) becomes increasingly prevalent in distribution networks, it can be used to quickly restore power to affected areas after a fault by isolating them in islanded configurations, ensuring power supply to critical loads and improving system reliability. Existing technologies for distribution networks containing DG reduce the scope of power outages and minimize losses by isolating areas and reconfiguring networks after a fault, while ensuring continuous power supply to critical loads.

[0004] Ensuring the restoration and continuous supply of power to critical loads after a distribution network fault is of paramount importance. Current fault recovery strategies often employ islanding methods that focus on distributed generation sources, with less emphasis on ensuring power supply to critical loads. Furthermore, these strategies frequently create islands powered by a single source, resulting in weak robustness, and fail to consider the interconnections between islands to facilitate power exchange and improve power reliability. Summary of the Invention

[0005] Based on the above analysis, the embodiments of the present invention aim to provide a distribution network fault recovery method and system based on load-side islanding. By dividing islands starting from primary loads, considering the power supply redundancy of each island to establish connections between islands, and considering the synergistic effect of multiple types of power sources, the method effectively ensures the path connectivity of power sources to important loads, prioritizes and quickly restores the power supply to important loads and ensures their continuous power supply, increases the coupling connection between islands, and improves the reliability, robustness, and resistance to damage of power supply.

[0006] This invention provides a distribution network fault recovery method based on load-side islanding, comprising the following steps: Step S1: Based on the distribution network topology, obtain the connectivity graph and adjust the edge weights. Starting from the primary load, search for the initial islands. Using the initial islands as the search starting point, use the edge weights to search for the remaining loads to be assigned to the islands, and obtain the distribution network island division results and its objective function. Step S2: Calculate the power supply redundancy of each island in the island division result, perform island fusion based on the power supply redundancy of each island to obtain the island fusion result, and obtain the objective function of the island fusion result based on the objective function of the island division result. Step S3: When a distribution network fault occurs, based on the islanding fusion results, its objective function, and constraints, the maximum load restoration amount of the distribution network is obtained to restore power supply to the load.

[0007] Further, step S1 includes: A connectivity graph is obtained based on the distribution network topology and distribution network system parameters; The edge weights of the connected graph are tuned and sorted in ascending order. Starting from each primary load, search for the distributed power source with the smallest edge weight. Each primary load and the corresponding distributed power source with the smallest edge weight form multiple initial islands. If it is determined that there are overlapping distributed power sources among the initial islands, they are merged. When it is determined that the merged islands satisfy the island partitioning constraints, multiple initial islands are obtained. Starting from the search origin, the remaining load is searched using edge weights and assigned to islands, resulting in multiple optimized islands. These optimized islands then form the island partitioning result. Based on the island partitioning results, an objective function for the island partitioning results is established.

[0008] Furthermore, the process of searching for remaining loads and assigning them to islands using edge weights from the search starting point yields multiple optimized islands, including: Each of the aforementioned initial islands serves as the starting point for the search; Starting from the search starting point, the search proceeds in ascending order of edge weights. If the initial island is not fully loaded and does not form a loop, then loads that meet the island partitioning constraints, as well as loads that are close to the island and of high importance, are assigned to the initial island. After traversing all connected graphs, if there are still loads that have not been included in the initial islands, then these loads are removed, resulting in multiple optimized islands.

[0009] Furthermore, the objective function of the island partitioning result aims to maximize the load restoration value within the islands, including:

[0010] in, Let the load recovery value function be... The set of nodes within the isolated island; For nodes The load weighting coefficient; For nodes Total load active power; For integer variables, Indicates power loss load On the isolated island, restore power; otherwise... .

[0011] Furthermore, the island merging result obtained by performing island merging based on the power supply redundancy of each island includes: Based on the power supply redundancy of each island, estimate the island operating frequency and node voltage at both ends of the tie line; Determine whether the islanding operating frequency and node voltage at both ends of the tie line meet the islanding merging constraint and load regulation constraint. If they do, the tie line between the two islands can be closed. Based on the connection lines between closable isolated islands, island merging is performed to obtain the island merging result.

[0012] Furthermore, the objective function for the island consolidation result, based on the objective function for the island partitioning result, establishes a load restoration quantity function to ensure that the load restoration power supply range is sufficiently large, while also considering network losses during fault recovery, including:

[0013] in, and These are the load recovery function and the network loss function, respectively. Indicates being on an isolated island Internal load active power, Let i be the load weighting coefficient for node i. This indicates the duration of each stage in the recovery process; Represents the set of all stages; Integer variables that are either 0 or 1 Representative load On the deserted island Power has been restored in some areas, while it has not been restored in others. ; This represents the network loss of the entire system.

[0014] Furthermore, the constraints include power frequency constraints and distribution network operation constraints; the distribution network operation constraints include distributed power output constraints, energy storage operation constraints, distribution network topology constraints, system power flow constraints, and system security constraints. The system safety constraints include operating voltage constraints and branch capacity constraints.

[0015] Furthermore, obtaining the maximum load restoration amount of the distribution network includes: Based on the objective function and constraints of the island fusion result, the operation constraints of the power distribution network are relaxed. The distribution network topology constraints, system power flow constraints, and system security constraints are transformed into a mixed-integer second-order cone programming model by applying rotating cone constraints. The mixed-integer second-order cone programming model is solved based on island fusion constraints and load regulation constraints. At the same time, based on the island fusion results, islands with high power supply redundancy are preferentially merged with islands with low power supply redundancy to obtain the maximum load recovery of the distribution network.

[0016] Furthermore, the loads are classified into primary loads, secondary loads, and tertiary loads according to their importance, with primary loads and secondary loads being the most important loads. When restoring power, ensure that all primary loads are restored and that secondary loads are restored to the greatest extent possible.

[0017] This invention also provides a distribution network fault recovery system based on load-side islanding, characterized in that it includes: The islanding module M1 is used to obtain a connectivity graph based on the distribution network topology and adjust the edge weights. Starting from the primary load, it searches for the initial islands. Using the initial islands as the search starting point, it searches for the remaining loads to be assigned to the islands using the edge weights, and obtains the distribution network islanding results and its objective function. The island fusion module M2 is used to calculate the power supply redundancy of each island in the island division result, perform island fusion based on the power supply redundancy of each island to obtain the island fusion result, and obtain the objective function of the island fusion result based on the objective function of the island division result. The distribution network load power restoration module M3 is used to obtain the maximum load restoration amount of the distribution network based on the island fusion results, its objective function, and constraints when the distribution network fails, so as to restore power supply to the load.

[0018] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects: 1. This invention divides the system into islands starting from the primary load, prioritizing the rapid restoration and continuous power supply to critical loads, and effectively ensuring the connectivity of the path from the source to the critical load. It fully considers the load side, ensuring 100% power supply to primary loads while maximizing the value of load restoration. 2. This invention improves the Kruskal algorithm by changing its edge weight tuning principle to make it applicable to solving the island partitioning model based on the load side. It searches for the minimum spanning tree starting from the first-level load to obtain the optimal island partitioning scheme; 3. In the fault recovery phase, this invention proposes island fusion based on power supply redundancy. By combining high-redundancy islands with low-redundancy islands, the power supply reliability of low-redundancy islands is improved, enabling power exchange between islands. 4. This invention effectively reduces the number of single-power islands obtained by solving the islanding model, enhances the power coupling relationship between islands, improves the robustness of islands, helps resist the negative impact of distributed power source failure and load fluctuation on load power supply, ensures reliable power supply for important loads, and improves the flexibility of the power distribution system. 5. This invention considers the synergistic effect of distributed power sources and energy storage, as well as load regulation, increasing the possible forms of island consolidation. Simultaneously, it considers power supply redundancy to more intuitively enhance the robustness of low-redundancy islands. During the islanding phase, restoring loads is considered to have the highest value, with 100% power restoration for primary loads. In the fault recovery phase, island consolidation is considered, enhancing the system's power supply reliability during fault recovery and improving the utilization rate of power generation resources.

[0019] In this invention, the above-described technical solutions can be combined with each other to achieve more preferred combinations. Other features and advantages of this invention will be set forth in the following description, and some advantages may become apparent from the description or be learned by practicing the invention. The objects and other advantages of this invention can be realized and obtained from what is particularly pointed out in the description and drawings. Attached Figure Description

[0020] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts.

[0021] Figure 1a This is a schematic diagram of the distribution network fault recovery method based on load-side islanding. Figure 1b A detailed schematic diagram of the distribution network fault recovery method based on load-side islanding; Figure 2 A flowchart illustrating the process of determining the starting point of the search; Figure 3 A schematic diagram illustrating the process of dividing an island into separate sections; Figure 4 A schematic diagram of the islanding merging process considering power supply redundancy; Figure 5 To improve the schematic diagram of the PG&E69 node power distribution system; Figure 6 This is a schematic diagram of a power distribution network fault. Figure 7a This is a schematic diagram of a power distribution network fault recovery strategy. Figure 7b This is a schematic diagram showing the results of the island division. Figure 7c This is a schematic diagram of the distribution network fault recovery strategy for comparison strategy 1; Figure 7d This is a schematic diagram of the distribution network fault recovery strategy for comparison strategy 2; Figure 8 This is a schematic diagram of a distribution network fault recovery system based on load-side islanding. Detailed Implementation

[0022] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

[0023] The following will clearly and completely describe the concept, specific structure, and technical effects of the present invention in conjunction with embodiments and accompanying drawings, so as to fully understand the purpose, features, and effects of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are all within the scope of protection of the present invention.

[0024] The distribution network fault recovery method and system provided in this solution based on load-side islanding divides the network into islands, which is currently less considered. At the same time, during the fault recovery period, island power supply redundancy is taken into account to merge the islands. The purpose of combining the two is to fully consider the needs of the load side, ensure the continuous power supply of important loads, establish the connection between islands to reduce the number of individual power sources, improve power supply reliability, and enhance the islands' immunity to disturbances.

[0025] Example 1: The distribution network fault recovery method based on load-side islanding provided in this solution, such as Figure 1a As shown, steps S1-S3 are included: Step S1: Based on the distribution network topology, obtain the connectivity graph and adjust the edge weights. Starting from the primary load, search for the initial islands. Using the initial islands as the search starting point, use the edge weights to search for the remaining loads to be assigned to the islands, and obtain the distribution network island division results and its objective function. Step S2: Calculate the power supply redundancy of each island in the island division result, perform island fusion based on the power supply redundancy of each island to obtain the island fusion result, and obtain the objective function of the island fusion result based on the objective function of the island division result. Step S3: When a distribution network fault occurs, based on the islanding fusion results, its objective function, and constraints, the maximum load restoration amount of the distribution network is obtained to restore power supply to the load.

[0026] The classification of the importance of distribution network loads in this plan includes: Level 1 load: A load for which a power outage would pose a risk of personal injury or death, or cause serious and irreparable damage to equipment, or cause significant political and economic losses. Secondary loads: Loads that would result in a large amount of scrap, material waste, significant production reduction, or major equipment damage if power outages were to occur, but which can be avoided by taking appropriate measures. Level 3 loads: General loads other than Level 1 and Level 2 loads. The losses caused by short-term power outages of Level 3 loads are not significant, such as factory workshops, small towns, and rural residential electricity use.

[0027] When restoring power, ensure that all primary loads are restored and that secondary loads are restored to the greatest extent possible. In this plan, important loads include primary and secondary loads, and the restoration plan should try to ensure the power supply to important loads.

[0028] This scheme first uses an improved Kruskal algorithm to solve the island partitioning model, and then searches for the minimum spanning tree to obtain the optimal island partitioning scheme that satisfies the island partitioning constraints. See the attached flowchart for details. Figure 3 Based on the islanding results, a network reconfiguration scheme is obtained by optimizing the islanded converged power distribution system recovery decision model containing integer variables and nonlinear constraints. See the attached flowchart for details. Figure 4 This invention's distribution network fault recovery method starts with load-based islanding and considers island merging during fault recovery, resulting in a method that ensures continuous power supply to critical loads, establishes connections between islands, and improves power supply reliability, power generation resource utilization, and load recovery rate. See the attached flowchart for a detailed flowchart. Figure 1b .

[0029] Step S1, specifically.

[0030] This invention first employs the Kruskal algorithm to solve the islanding model of a distribution network considering multiple islands from the load side, including steps S11-S15: Step S11: Obtain the connectivity graph based on the distribution network topology and distribution network system parameters.

[0031] Based on the distribution network model topology, a connected graph corresponding to the distribution network topology is generated; the connected graph is an undirected graph abstracted from the tie lines in the topology graph of the distribution network system by assigning weights.

[0032] The vertex types in a connected graph include: busbars, distributed generation, and loads; Edge types are divided into three categories based on vertex types: busbar to busbar, busbar to load, and busbar to distributed power source.

[0033] Step S12: Adjust the edge weights of the connected graph and sort the edge weights in ascending order.

[0034] Based on the connected graph, the Kruskal algorithm is used to tune the edge weights and sort them in ascending order according to the edge weights. The rule for tuning edge weights in a connected graph using Kruskal's algorithm is to traverse the fully connected graph and search for the minimum spanning tree formed by the combination of edges with the minimum weights that do not constitute cycles. The edge weight tuning is as follows: ① The edges whose vertices are all composed of busbars. Since the island division principle considers minimizing network loss within the island in order to maximize load, the resetting of the edge weight depends on whether there is a transformer. If there is no transformer, the network loss is smaller and should be given priority, so the weight should be set to be smaller. Otherwise, it is larger. Therefore, the weight of the edge without transformer is set to 1, and the weight of the edge with transformer is set to 2. ② The vertex is the edge formed by the bus and the distributed power source. Since the distributed power source with the shortest distance is searched first when dividing the island, the initial island is the search starting point of the algorithm. Therefore, the weight of this type of edge is set to 0. ③ The vertex is the edge formed by the busbar and the load. Since the importance of the load should be considered when dividing the island, important loads should be kept powered as much as possible. Therefore, the weights of the first, second and third level loads are set to 1, 2 and 3, respectively. At the same time, since the distributed power source with the shortest distance is found first-level load when dividing the island, the distance between the load and the busbar and the size of the load should be considered to ensure the maximum load recovery. Therefore, the edge weights are set as shown in formula (1): (1) in, Indicates that the vertex is a load. The edge weight of the edge connected to the bus. , , These represent the proportions of load importance, load size, and load distance from the busbar when considering all factors. They are constants and are set according to actual conditions. Indicates load The importance level has been adjusted to the load weight. . It indicates the distance of the load from the busbar, expressed as line impedance. Represents the load after standardization The size of is shown in formula (2).

[0035] (2) in, , , Representing loads respectively Size, minimum value among all loads in the system, maximum value among all loads, and all of them are active power.

[0036] Sort the edge weights in ascending order.

[0037] Step S13: Starting from the primary load, search for the distributed power source with the smallest edge weight. The primary load and the distributed power source form multiple initial islands. If there is overlap between distributed power sources in the initial islands, they are merged. Determine whether the merged islands meet the island partitioning constraints to obtain multiple initial islands.

[0038] The first step is to find the nearest distributed power source starting from the primary load. The primary load and the distributed power source form multiple initial islands. Among them, the closest one is the one with the smallest edge weight.

[0039] The second step is to determine whether there is any overlap in distributed power sources among the initial islands. If there is overlap in distributed power sources among the initial islands, they are merged. The merged islands are then judged to meet the island partitioning constraints, thus obtaining multiple initial islands.

[0040] For each initial isolated island, determine whether there is any overlap in distributed power sources between them, in the following cases: ① Determine whether there is any overlap in distributed power sources between the initial islands. If there is no overlap in distributed power sources, then the initial island is an initialization island; Otherwise, if there is overlap of distributed power sources, merge the initial islands and determine whether the merged islands meet the island partitioning constraints. If they meet the power constraints within the islands, they are merged into the initial islands. If the merged islands cannot meet the islanding constraints, the primary load with the largest load and the overlapping distributed power source will be assigned to the merged islands. The remaining primary loads, after removing the primary load with the largest load, will be re-searched for the nearest original island and assigned to that original island.

[0041] Remove the other distributed power sources that overlap with the original distributed power source, and the original island is re-formed.

[0042] ②If a separate distributed power source exists, it forms an initialization island.

[0043] The initial island was determined as the starting point for the search.

[0044] See the attached flowchart for the detailed process. Figure 2 .

[0045] Among them, islanding constraints include: power balance constraints within islands and transmission line security constraints within islands.

[0046] As a small power system, an island must achieve a power balance between the power generated by the power source and the power demand of the load in order to operate stably. Therefore, it is necessary to establish power balance constraints within the island. Furthermore, since there is an upper limit to the power flowing through the transmission lines, safety constraints for the transmission lines within the island are also established.

[0047] Establish power balance constraints within the island, as shown in formulas (3)-(7).

[0048] (3) (4) (5) (6) (7) in, For nodes Electrical energy stored and discharged; For energy storage capacity; For integer variables, Indicates power loss load On the isolated island, restore power; otherwise... ; The state of charge of energy storage; , , These are photovoltaic, wind power, and load, respectively. Active power at any given moment; , , During the fault period The electricity provided by nodal photovoltaic and wind power, as well as the electricity consumed by the load. The moment the fault begins. This is the time when the fault ends.

[0049] Establish security constraints for transmission lines within the island, as shown in formula (8).

[0050] (8) in, For the collection of distribution network branches, Indicates the line flowing through the island power, Indicates the line The maximum power transmission capacity.

[0051] This step yields an initial island, which serves as the starting point for the search in step S14.

[0052] Step S14: Starting from the search starting point, use edge weights to search for the remaining load to be assigned to the island, resulting in multiple optimized islands. The optimized islands constitute the island partitioning result.

[0053] The first step is to start from the search starting point and search according to the edge weights from smallest to largest. If the initial island is not fully loaded and does not form a loop, then loads that meet the island partitioning constraints are assigned to the initial island based on their distance from the distributed power sources at each search starting point and the importance of the loads. Loads that are close to the source and have high importance are assigned to that initial island. "Close to the source" refers to loads with small edge weights; "high importance" refers to loads that are first-level or second-level loads.

[0054] After traversing the connected graph, ensure that the load of each initial island is maximized. If there is an initial island that is fully loaded and satisfies the island partitioning constraints, then assign that load to another initial island according to the principle of proximity.

[0055] The second step is to establish a minimum spanning tree. After traversing all edges of the connected graph, if there are still loads that have not been included in the initial islands, then remove such loads, optimize the spanning tree, obtain the minimum spanning tree, and obtain multiple optimized islands. Multiple optimized islands form the island partitioning model of the distribution network from the load side, which is the optimal island partitioning scheme.

[0056] This step involves initializing multiple isolated islands and then searching them using an improved Kruskal algorithm to obtain multiple optimized islands. The reason it's called an improved Kruskal algorithm is that existing Kruskal algorithms don't fix the search starting point, which can easily lead to non-unique island optimization results. This step, by fixing the search starting point, ensures a unique and optimal island partitioning scheme.

[0057] Step S15: Based on the island partitioning results, establish the objective function for the island partitioning results.

[0058] Based on the islanding model of the distribution network considering the load side, the objective function is to maximize the load restoration value within the island, as shown in formula (9).

[0059] (9) in, Let the load recovery value function be... The set of nodes within the isolated island; For nodes The load weighting coefficient; For nodes Total load active power; For integer variables, Indicates power loss load On the isolated island, restore power; otherwise... .

[0060] Step S1 involves dividing the distribution network into islands to obtain multiple optimized islands, reducing the number of single-source islands and laying the foundation for island merging in the fault recovery phase of step S2.

[0061] Step S2, specifically.

[0062] The islanding merging method considering power supply redundancy establishes connections between islands during fault recovery based on the optimal islanding partitioning result, taking into account the power supply redundancy of each island. This includes steps S21-S22: Step S21: Calculate the power supply redundancy of each island in the islanding results.

[0063] Establish an island power supply redundancy function and calculate the power supply redundancy of each island in the island partitioning model, as shown in formula (10): (10) in, Indicates an isolated island The power supply redundancy represents the redundancy margin of power supply within the island; Indicates an isolated island The sum of active power that can be generated by distributed power sources within the area; Indicates an isolated island The sum of the energy storage charging and discharging power, with charging power being positive and discharging power being negative; Indicates being designated as an isolated island The sum of active power absorbed by the load within the area; It represents the total active power that can be generated by all distributed power sources in the distribution network.

[0064] The larger the size, the more isolated the island. The more power the internal power supply can provide, the more priority it should be during fault recovery. Large number of isolated islands and Small-value islands are merged to ensure that as many loads as possible can be restored in the event of distributed power source failure or load fluctuation. Power can be transferred between islands, serving as backups for each other and improving power supply reliability.

[0065] Step S22: Based on the power supply redundancy of each island, island fusion is performed to obtain island fusion result, and based on the objective function of the island division result, the objective function of island fusion result is obtained.

[0066] The first step is to estimate the operating frequency and node voltage of the islands at both ends of the tie line based on the power supply redundancy of each island.

[0067] When the power supply redundancy of an island is high enough, the island has sufficient capacity to meet load demands, which helps maintain frequency and voltage stability.

[0068] Insufficient power supply redundancy, especially under conditions of rapid load increase or insufficient power supply capacity, leads to power reduction and voltage fluctuations. This affects the stability of the island and the normal operation of equipment. Tie lines are lines connecting different islands. Based on the power supply redundancy of each island, and using existing island power system models, the operating frequencies and node voltages of the islands at both ends of the tie lines are estimated.

[0069] The second step is to determine whether the islanding operating frequency and node voltage at both ends of the tie line meet the islanding merging constraint and load regulation constraint. If they do, the tie line between the two islands can be closed.

[0070] The island fusion constraints for the island fusion model are established as shown in formulas (11)-(12): (11) (12) in, Represents a node The square of the voltage amplitude; Represents a node The square of the voltage amplitude; This represents the voltage coefficient, with a value of 2. Indicates an isolated island Frequency within, Indicates an isolated island The frequency within. , These represent the allowable voltage deviation of the loads at both ends of the tie line and the allowable frequency deviation of the island, respectively. Since the reference voltage is 12.66kV, This is a per-unit value, taking the value 0.001. The value is 0.5Hz. , These represent the set of connecting lines and the set of isolated islands, respectively.

[0071] Determine whether the island operating frequency and node voltage at both ends of the tie line meet the island merging constraint. If the island merging constraint is met, determine the number of islands that meet the island merging constraint conditions, determine the tie lines between islands that can be closed, and prioritize powering islands with large differences in island redundancy.

[0072] If the islanding merging constraint is not met, it is determined whether there are unrestored or incompletely restored loads within the island. If so, islanding merging constraints and load control constraints are used to selectively restore the loads. Then, the number of restored loads and the number of newly added islands that meet the merging constraint are calculated. Closeable interconnections between islands are determined, and islands with large differences in redundancy are prioritized for power supply. Otherwise, the islands do not participate in the islanding merging process.

[0073] Establish load regulation constraints. When the island fusion constraint conditions are not met, selectively restore unrestored or controllable loads so that more islands meet the island fusion constraints. The load regulation constraints are shown in formula (13): (13) Equation (13) represents the range within which all unrestored or controllable loads can be adjusted, where Indicates the number of loads undergoing load adjustment. This represents a set of loads that have not been restored or are controllable. Indicates all All loads within the area are connected.

[0074] Establish a controllable load model, as shown in formulas (14)-(15).

[0075] (14) (15) in, Represents a node The active power of the load, For nodes The active power of uncontrollable loads; For nodes The active power of the controllable load; For nodes The maximum active power of the controllable load.

[0076] The third step is to perform island fusion based on the connection lines between closable islands, and obtain the island fusion result and objective function.

[0077] Based on the established interconnections between isolated islands, island merging is performed, and the remaining unrestored loads are then restored, resulting in an island merging model. This model contains multiple optimized merged islands, which are then restored during distribution network fault recovery.

[0078] Establish the objective function of the island fusion model, and establish the load restoration quantity function based on the objective function of the island division result to ensure that the load restoration power supply range is large enough, while considering the network loss during the fault restoration process, as shown in formula (16): (16) in, and These are the load recovery function and the network loss function, respectively. Indicates being on an isolated island Internal load active power, Let i be the load weighting coefficient for node i. This indicates the duration of each stage in the recovery process; Represents the set of all stages; Integer variables that are either 0 or 1 Representative load On the deserted island Power has been restored in some areas, while it has not been restored in others. ; This represents the network loss of the entire system.

[0079] Compared to the single-power-source islanding of the islanding model obtained in step S1, this islanding fusion model considers power supply redundancy for optimized fusion, which is beneficial for restoring faulty loads during distribution network fault recovery.

[0080] Step S3, specifically.

[0081] It includes the following steps.

[0082] Step S31: Construct constraints.

[0083] The constraints include power frequency constraints and distribution network operation constraints; the distribution network operation constraints include distributed power output constraints, energy storage operation constraints, distribution network topology constraints, system power flow constraints, and system security constraints; wherein, the system security constraints include operating voltage constraints and branch capacity constraints.

[0084] The first step is to establish power supply frequency constraints, as shown in formulas (17)-(20): (17) (18) (19) (20) in, , These are the upper and lower limits of the power supply frequency, respectively, taken as 49.5Hz and 50.5Hz; The power droop factor is approximated by the ratio of the allowable frequency fluctuation range of each power source to its maximum output. For the isolated island Distributed power sources within Maximum active power output; For power supply During the period Internal power supply frequency; Indicates an isolated island Internal frequency.

[0085] The second step is to establish distribution network operation constraints, including distributed generation output constraints, energy storage operation constraints, distribution network topology constraints, system power flow constraints, and system security constraints. Among these, system security constraints include operating voltage constraints and branch capacity constraints.

[0086] Establish the output constraints of the distributed power source, as shown in formulas (21)-(23): (twenty one) (twenty two) (twenty three) in, , Time periods Inside The active and reactive power generated by the distributed power sources at the nodes; , Time period Inside The upper and lower limits of the active power output of the distributed power source at the node; for The capacity of the distributed power source connected to the node; For nodes Minimum operating power factor of distributed power sources at the location; Establish energy storage operation constraints, as shown in formulas (24)-(28).

[0087] By utilizing different energy storage outputs, more isolated islands can meet the constraints of island integration, thus expanding the scope of island integration: (twenty four) (25) (26) (27) (28) in, Time period Inside The energy storage charging and discharging power at the node is as follows: charging power is positive, and discharging power is negative. Time period Inside The reactive power generated by the energy storage at the node; for The energy storage capacity at the node; for The upper limit of reactive power that can be generated by energy storage at the node. express The upper limit of reactive power that energy storage at the node can absorb; Time period Initial time Energy storage charge state at the node; The time interval step; To start from the initial time period Time period The state of charge of the energy stored at node j is Time period internal nodes Energy storage operation losses; For access nodes Loss coefficient of the upper energy storage system; and They are nodes The upper and lower limits of the state of charge of the energy storage.

[0088] Establish distribution network topology constraints to ensure the radial network structure of the distribution network, as shown in formulas (29)-(31): (29) (30) (31) in, For distribution network branch collection; The set of nodes that have lost power; Indicates a branch The on / off state, This indicates that the switch is closed. This indicates that the switch is off; and Representative node and In a parent-child relationship, if node... It is a node The parent node, then and Conversely and If node With nodes If not connected, then .

[0089] Establish system power flow constraints, as shown in formulas (32)-(37). When a fault occurs in the distribution network, the network topology changes. Improve the traditional Distflow power flow model by introducing line disconnection variables. By relaxing the power flow equations, we obtain Distflow power flow operation constraints applicable to distribution network fault recovery.

[0090] (32) (33) (34) (35) (36) (37) in, , All are line opening / closing variables. Indicates the line Time period The internal communication switch is closed. Indicates the line Time period The internal communication switch is closed. Indicates the line Time period The internal communication switch is off. Indicates the line Time period The internal communication switch is off; for Time Branch By node Flow to Node The active power; Time period internal nodes Flow to Node The current amplitude; Time period internal nodes The voltage amplitude; branch road The resistance; branch road The reactance; for Time-of-day branch By node Flow to Node reactive power; , They are respectively Time period branch Transmitted active and reactive power; , Time periods internal nodes Injected active and reactive power; , Time periods Internal Distributed Power Injection Node Active power and reactive power; , Time periods internal nodes The active and reactive power released from the energy storage at the location; , They are respectively Time period nodes The active and reactive power consumed by the load; Time period internal nodes The square of the voltage amplitude; For line opening and closing variables, Indicates the line Time period The internal communication switch is closed. Indicates the line Time period When the internal connecting switch is open, the line switch variables in equations (36) and (37) are... This equation constraint holds only for closed branches; and These are nodes in the distribution network. The set of parent and child nodes.

[0091] System safety constraints include operating voltage constraints and branch capacity constraints.

[0092] Establish operating voltage constraints, as shown in formula (38): (38) in, and They are nodes The upper and lower limits of voltage.

[0093] Establish branch capacity constraints, as shown in formula (39): (39) in, For the collection of distribution network branches, branch road The upper limit of the current.

[0094] Step S32: Based on the objective function and constraints of the island fusion result, the operation constraints of the distribution network are relaxed to obtain the maximum load restoration amount of the distribution network, so as to restore power supply to the load.

[0095] The islanding consolidation model considering islanded power supply redundancy contains integer variables and nonlinear constraints. Relaxing the nonlinear constraints transforms them into convex constraints, and the entire model is transformed into a mixed-integer second-order cone programming model, which belongs to convex optimization. Due to the limitations of the constraints, transforming it into a second-order cone model facilitates solving the objective function of the islanding consolidation model.

[0096] Second-order cone programming (SOCP) is an important mathematical model and optimization problem in the fields of mathematical programming and optimization. It is a convex optimization problem used to find the optimal values ​​of a set of variables under a set of constraints, so as to maximize or minimize the value of a linear function.

[0097] The first step is to relax the distribution network operation constraints based on the objective function and constraints of the island fusion result; and to transform the distribution network topology constraints, system power flow constraints and system security constraints into a mixed integer second-order cone programming model by rotating cone constraints.

[0098] Based on the objective function and constraints of the island fusion results, the distribution network operation constraints are relaxed. These constraints include distributed generation output constraints, energy storage operation constraints, distribution network topology constraints, system power flow constraints, and system security constraints, including operating voltage constraints and branch capacity constraints, which are then relaxed.

[0099] A second-order cone model is used to transform the constraints in the distribution network topology constraints, system power flow constraints, and system security constraints within the operational constraints of the distribution network into a mixed-integer second-order cone model through rotating cone constraints. This includes: The constraints are transformed using a second-order cone model. The constraints (23) of the distributed power output constraint and (24) and (26) of the energy storage operation constraint are transformed into a mixed integer second-order cone programming model by rotating cone constraints, as shown in formulas (41)-(42).

[0100] (40) (41) (42) Further transform equation (42) into a second-order cone constraint, as shown in equation (43).

[0101] (43) The second step is to solve the mixed-integer second-order cone programming model, including island merging constraints and load regulation constraints. At the same time, based on the island merging results, priority is given to merging islands with high power supply redundancy with islands with low power supply redundancy to obtain the maximum load recovery amount considering power supply redundancy and island merging.

[0102] (1) Using YALMIP programming, construct a mixed integer second-order cone programming model in MATLAB using the YALMIP toolbox, and import the objective function and constraints of the islanded mixed model.

[0103] For the aforementioned system power flow constraints, equations (36) and (37) are subject to the following conditions. This approach is only suitable for closed branches. To broaden its applicability to the entire energy network, rather than just closed branches, a convex relaxation process is employed, introducing variables. and replace and By introducing inequality constraints to relax the operation constraints and topology constraints of the distribution network, the active power, reactive power and line current of the disconnected branch are zero, and there are no constraints on the closed branch, as shown in formulas (44)-(51): (44) (45) (46) (47) (48) (49) (50) (51) in, , , and It is a sufficiently large positive number; For line opening and closing variables, Indicates the line Time period The internal communication switch is closed. Indicates the line Time period The internal communication switch is off; Time period internal nodes The square of the voltage amplitude; for Time period branch By node Flow to Node The active power; for Time period branch By node Flow to Node reactive power; Time period internal nodes Flow to Node The square of the current amplitude; Time period internal nodes The square of the voltage amplitude; branch road The resistance; branch road The reactance; , They are respectively Time Branch Transmitted active and reactive power; , Time periods internal nodes Injected active and reactive power; and These are nodes in the distribution network. The set of parent and child nodes; This is a collection of distribution network branches.

[0104] (2) Select a solver. Choose a suitable mathematical optimization solver, such as CPLEX or MOSEK, to solve the mixed integer second-order cone programming model. (3) Solve the model using the selected solver to obtain the optimal solution or an approximate optimal solution; (4) Island fusion strategy: Based on the solution results, priority is given to merging islands with high power supply redundancy with islands with low power supply redundancy to improve the reliability and recovery capability of the distribution network system; Consider load regulation and load regulation constraints, and adjust the load as needed to meet the needs of the distribution network.

[0105] (5) Evaluation results, adjustment and optimization: evaluate the solution results of the model to ensure that all constraints are met. Based on the evaluation results, adjust and optimize the islanding fusion model to further improve the distribution network recovery strategy.

[0106] After the above steps, the fault recovery decision model of the islanded integrated distribution system with integer variables and nonlinear constraints, which considers power supply redundancy, is transformed from a mixed integer nonlinear power supply recovery model that is difficult to solve efficiently into a second-order cone programming model. Using YAMIP programming and CPLEX and MOSEK software, a distribution network recovery strategy based on load-side islanding and considering power supply redundancy during fault recovery is obtained.

[0107] This invention provides a distribution network fault recovery method based on load-side islanding. To verify the effectiveness of this method, a specific example is provided: the improved PG&E69 node distribution system is shown in the attached figure. Figure 5 As shown, this PG&E69 node distribution system incorporates distributed generation, specifically including two wind power plants, two photovoltaic power plants, and two energy storage devices. Branch impedance, load size, and load level parameters are shown in Table 1. Node load weight levels, load weight coefficients, and load setting weights are shown in Table 2. The weights for level 1, 2, and 3 loads are 100, 10, and 1, respectively, and the setting weights for level 1, 2, and 3 loads are 1, 2, and 3, respectively. Distributed generation parameters are shown in Table 3, and energy storage parameters are shown in Table 4.

[0108] Table 1 Parameters of the PG&69 Node Power Distribution System

[0109] Table 2 Node Load Levels

[0110] Table 3 Distributed power generation parameters

[0111] Table 4 Energy Storage Parameters

[0112] The fault conditions of the improved PG&E69 node matching system are attached. Figure 6 As shown, if there are 7 faults in the line and the faults cannot be repaired in a short time, it is considered a broken line.

[0113] The strategy of this invention is to use the distribution network fault recovery method based on load-side islanding proposed in this invention. When solving the islanding model, the optimal islands for the islanding model are generated starting from the load side. Based on the aforementioned islanding results, island fusion, taking into account the redundancy of island power supply, is performed during the fault recovery phase to obtain the final distribution network fault recovery strategy. See Figure 1 for the specific flowchart.

[0114] Comparison Strategy 1: Islanding is performed from the load side as proposed in this invention, but power supply redundancy is not considered during island merging during fault recovery. (Comparison Strategy 1 Appendix) Figure 7c The solution considers island consolidation but does not take power supply redundancy into account. The aim is to compare the results with the proposed solution to highlight the improved distributed power supply utilization efficiency of the proposed solution that takes island power supply redundancy into account, and to further enhance the range and quantity of power supply restoration for the load.

[0115] Comparison Strategy 2: The island partitioning method proposed in this invention is not used, and the island merging process is not considered during fault recovery.

[0116] Comparison Strategy 2 (Appendix) Figure 7d Neither using a load-side-oriented partitioning scheme nor considering an island merging model, this approach aims to complement the island partitioning results presented in this paper. Figure 7b The comparison shows that the islanding model in this scheme is more beneficial for critical load recovery, compared with the attached scheme. Figure 7a The comparison demonstrates that this solution is more beneficial for power utilization and load recovery.

[0117] Using the distribution network fault recovery method based on load-side islanding proposed in this invention, the fault recovery scenario obtained is as follows: Figure 7a As shown, three islands are ultimately formed, all primary loads are restored, and the load restoration area is relatively large. The islanding scheme obtained using the distribution network fault recovery method based on load-side islanding proposed in this invention is illustrated in the diagram below. Figure 7b As shown, the result is four isolated islands, and the range of load restoration power supply is relatively small; The fault recovery scenario obtained by comparing strategy 1 is as follows: Figure 7c As shown, three islands were obtained, but the secondary load 25 failed to be restored, which is slightly worse than the restoration of important loads in the present invention. A schematic diagram of the island partitioning scheme obtained by comparison strategy 2 is shown below. Figure 7d As shown, four islands are formed. Compared with the island division result of the present invention, the secondary loads 3 and 28 are not restored, and the load restoration degree is not as good as that of the present invention.

[0118] Tables 5, 6, 7, and 8 show the power supply redundancy of each region in the island partitioning scheme obtained in this paper, the island partitioning scheme obtained in comparison strategy 2, the fused island formed by the recovery strategy in this paper, and the fused island formed by comparison strategy 1. The important loads restored by the island partitioning in this paper refer to the primary and secondary loads, as shown in Table 5, and the important loads restored by the island partitioning in comparison strategy 2 are shown in Table 6.

[0119] The power distribution network fault recovery scheme proposed in this invention, such as Figure 7aAs shown, while ensuring normal power supply to critical loads, the controllable load 50 is connected to the system to ensure that tie lines (50, 51) and (51, 52) meet the islanding merging conditions. From the data in the table and the schematic diagrams of each strategy scenario, it can be seen that: 1. Compare Tables 5 and 6 and Figure 7b , Figure 7d The islanding strategy of this invention provides higher power supply redundancy than the strategy without this invention. As shown in the diagram, although the islanding strategy without this invention improves the utilization of power generation resources, the islanding of islands 1 and 4 reduces the recovery of critical loads (13, 21) and (42, 43, 3, 28), thus reducing the value of load recovery. Ensuring the normal and continuous power supply to critical loads after a fault in the distribution network is of great significance for protecting people's lives and production, resisting disasters and accidents, and promoting social development. Therefore, the islanding strategy of this invention can guarantee the power supply to more critical loads.

[0120] 2. Compare Tables 5 and 7 and Figure 7b , Figure 7a It can be seen that, by using the fault recovery strategy of this invention, island merging that takes into account power supply redundancy based on load-based islanding can significantly reduce power supply redundancy during recovery, reducing it from 160.26 × 10⁻⁶. -3 Reduced to 88.37×10 -3 Furthermore, the power supply to the critical load 25 was restored, meaning that the fault recovery method of this invention can significantly improve the utilization rate of power generation resources, increase the load recovery rate, and improve the reliability of power supply.

[0121] 3. Compare Tables 7 and 8 and Figure 7a , Figure 7c It can be seen that the island consolidation scheme that takes into account the redundancy of islanded power supply improves the efficiency of power generation resource utilization and load recovery rate more than the island consolidation strategy that does not consider redundancy.

[0122] Table 5. Island power supply redundancy in this paper

[0123] Table 6 Comparison Strategy 2: Island Power Supply Redundancy

[0124] Table 7 Power supply redundancy in isolated areas according to the recovery strategy presented in this paper

[0125] Table 8 Comparison Strategy 1: Power Supply Redundancy in Isolated Areas

[0126] The above analysis leads to the following conclusions: (1) Using a load-based islanding strategy can improve the recovery rate of critical loads, ensure the recovery of critical loads, increase the value of load recovery, and safeguard people's lives and production. (2) Considering the island fusion process after islanding can improve the utilization rate of power generation resources, further improve the load recovery rate, reduce single-source power supply microgrids, and improve power supply reliability and anti-interference capability. (3) Island fusion that takes into account the redundancy of island power supply can improve the recovery rate of critical loads, ensure the power supply of critical loads, and significantly improve the utilization rate of power generation resources compared to the island fusion strategy that does not consider redundancy.

[0127] The distribution network fault recovery method based on load-side islanding provided by this invention divides islands from the load side and merges islands during fault recovery, taking into account the redundancy of island power supply. The combination of the two fully considers the needs of the load side, ensures the continuous power supply of important loads, establishes connections between islands to reduce the number of individual power sources, improves power supply reliability and island immunity, and improves the utilization rate of power generation resources.

[0128] Example 2: Another embodiment of the present invention discloses a distribution network fault recovery system based on load-side islanding, thereby implementing the distribution network fault recovery method based on load-side islanding in Embodiment 1. The specific implementation of each module is described in the corresponding description in Embodiment 1.

[0129] like Figure 8 As shown, the system includes an islanding module M1, an islanding fusion module M2, and a power distribution network load restoration module M3, which are respectively: The islanding module M1 is used to obtain a connectivity graph based on the distribution network topology and adjust the edge weights. Starting from the primary load, it searches for the initial islands. Using the initial islands as the search starting point, it searches for the remaining loads to be assigned to the islands using the edge weights, and obtains the distribution network islanding results and its objective function. The island fusion module M2 is used to calculate the power supply redundancy of each island in the island division result, perform island fusion based on the power supply redundancy of each island to obtain the island fusion result, and obtain the objective function of the island fusion result based on the objective function of the island division result. The distribution network load power restoration module M3 is used to obtain the maximum load restoration amount of the distribution network based on the island fusion results, its objective function, and constraints when the distribution network fails, so as to restore power supply to the load.

[0130] Since the system in this embodiment and the method in Embodiment 1 are related and can be referenced from each other, this description is redundant and will not be repeated here. Because this system embodiment shares the same principle as the above method embodiment, it also possesses the corresponding technical effects of the above method embodiment.

[0131] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware, and the program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0132] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A distribution network fault recovery method based on load-side islanding, characterized in that, Includes the following steps: Step S1: Based on the distribution network topology, obtain the connectivity graph and adjust the edge weights. Starting from the primary load, search for the initial islands. Using the initial islands as the search starting point, use the edge weights to search for the remaining loads to be assigned to the islands, and obtain the distribution network island division result and the objective function of the island division result. Step S2: Calculate the power supply redundancy of each island in the island division result, perform island fusion based on the power supply redundancy of each island to obtain the island fusion result, and obtain the objective function of the island fusion result based on the objective function of the island division result. Based on the islanding results, during fault recovery, connections between islands are established by considering the power supply redundancy of each island, including steps S21-S22: Step S21: Calculate the power supply redundancy of each island in the islanding results; Establish an island power supply redundancy function and calculate the power supply redundancy of each island in the island partitioning model, as shown in formula (1): Official (1) in, Indicates an isolated island Power supply redundancy; Indicates an isolated island The sum of active power that can be generated by distributed power sources within the area; Indicates an isolated island The sum of the energy storage charging and discharging power, with charging power being positive and discharging power being negative; Indicates being designated as an isolated island The sum of active power absorbed by the load within the area; It represents the total active power that can be generated by all distributed power sources in the distribution network; The set of nodes within the isolated island; Step S22: Based on the power supply redundancy of each island, island fusion is performed to obtain island fusion result, and based on the objective function of the island division result, the objective function of island fusion result is obtained; The first step is to calculate the island operating frequency and node voltage at both ends of the tie line based on the power supply redundancy of each island. The second step is to determine whether the islanding operating frequency and node voltage at both ends of the tie line meet the islanding merging constraint and load regulation constraint. If they do, it is determined that the tie line between the two islands can be closed. The island fusion constraints for the island fusion model are established as shown in formulas (2)-(3): Official (2) Official (3) in, Represents a node The square of the voltage amplitude; Represents a node The square of the voltage amplitude; This represents the voltage coefficient, with a value of 2. Indicates an isolated island Frequency within, Indicates an isolated island Frequency within; , These represent the allowable voltage deviation of the loads at both ends of the tie line and the allowable frequency deviation of the island, respectively; since the reference voltage is 12.66kV, This is a per-unit value, taking the value 0.

001. The value is 0.5Hz; , These represent the set of connecting lines and the set of isolated islands, respectively. If the island merging constraint is not met, it is determined whether there are unrestored or incompletely restored loads within the island. If so, island merging constraints and load control constraints are used to selectively restore the loads. Then, the number of restored loads and the number of newly added islands that meet the merging constraint are calculated. The interconnection lines between islands that can be closed are determined, and islands with large differences in power supply redundancy are given priority. Otherwise, the islands do not participate in the island merging process. A load control constraint is established. When the island fusion constraint is not met, selective restoration is performed on unrestored or controllable loads so that more islands meet the island fusion constraint. The load control constraint is shown in formula (4): Official (4) Formula (4) represents the range within which all unrestored or controllable loads can be adjusted, where Indicates the number of loads undergoing load adjustment. This represents a set of loads that have not been restored or are controllable. Indicates all All loads within the system are connected; Establish a controllable load model, as shown in formulas (5)-(6); Official (5) Official (6) in, Represents a node The active power of the load, For nodes The active power of uncontrollable loads; For nodes The active power of the controllable load; For nodes The maximum active power of the controllable load; The third step is to perform island fusion based on the connection lines between closable islands, and obtain the island fusion result and objective function. Based on the established interconnections between isolated islands, island merging is performed, and the remaining unrestored loads are restored, resulting in an island merging model. The island merging model contains multiple optimized merged islands, which are restored during distribution network fault recovery. Establish the objective function of the island fusion model, and establish the load restoration quantity function based on the objective function of the island division result to ensure that the load restoration power supply range is large enough, while considering the network loss during the fault restoration process, as shown in formula (7): Official (7) in, and These are the load unrecovery function and the network loss function, respectively. Indicates being on an isolated island Internal load active power, Let i be the load weighting coefficient for node i. This indicates the duration of each stage in the recovery process; Represents the set of all stages; Integer variables that are either 0 or 1 Representative load On the deserted island Power has been restored in some areas, while it has not been restored in others. ; Represents the network loss of the entire system; Step S3: When a distribution network fault occurs, based on the islanding fusion results, its objective function, and constraints, the maximum load restoration amount of the distribution network is obtained to restore power supply to the load.

2. The power distribution network fault recovery method according to claim 1, characterized in that, Step S1 includes: A connectivity graph is obtained based on the distribution network topology and distribution network system parameters; The edge weights of the connected graph are tuned and sorted in ascending order. Starting from each primary load, search for the distributed power source with the smallest edge weight. Each primary load and the corresponding distributed power source with the smallest edge weight form multiple initial islands. If it is determined that there are overlapping distributed power sources among the initial islands, they are merged. When it is determined that the merged islands satisfy the island partitioning constraints, multiple initial islands are obtained. Starting from the search origin, the remaining load is searched using edge weights and assigned to islands, resulting in multiple optimized islands. These optimized islands then form the island partitioning result. Based on the island partitioning results, an objective function for the island partitioning results is established.

3. The power distribution network fault recovery method according to claim 2, characterized in that, The process of searching for remaining loads and assigning them to islands using edge weights, starting from the search origin, yields multiple optimized islands, including: Each of the aforementioned initial islands serves as the starting point for the search; Starting from the search starting point, the search proceeds in ascending order of edge weights. If the initial island is not fully loaded and does not form a loop, then loads that meet the island partitioning constraints, as well as loads that are close to the island and of high importance, are assigned to the initial island. After traversing all connected graphs, if there are still loads that have not been included in the initial islands, then these loads are removed, resulting in multiple optimized islands.

4. The power distribution network fault recovery method according to claim 3, characterized in that, The objective function of the island partitioning result aims to maximize the load restoration value within each island, including: Official (8) in, This is a function representing the value of load recovery. For nodes The load weighting coefficient; For nodes Total load active power; For integer variables, Indicates power loss load On the isolated island, restore power; otherwise... .

5. The power distribution network fault recovery method according to claim 1, characterized in that, The constraints include power frequency constraints and distribution network operation constraints; the distribution network operation constraints include distributed power generation output constraints, energy storage operation constraints, distribution network topology constraints, system power flow constraints, and system security constraints. The system safety constraints include operating voltage constraints and branch capacity constraints.

6. The power distribution network fault recovery method according to claim 5, characterized in that, The process of obtaining the maximum load restoration amount of the distribution network includes: Based on the objective function and constraints of the island fusion result, the operation constraints of the power distribution network are relaxed. The distribution network topology constraints, system power flow constraints, and system security constraints are transformed into a mixed-integer second-order cone programming model by applying rotating cone constraints. The mixed-integer second-order cone programming model is solved based on island fusion constraints and load regulation constraints. At the same time, based on the island fusion results, islands with high power supply redundancy are preferentially merged with islands with low power supply redundancy to obtain the maximum load recovery of the distribution network.

7. The power distribution network fault recovery method according to any one of claims 1-6, characterized in that, Loads are classified into three levels according to their importance: Level 1 loads, Level 2 loads, and Level 3 loads. Among them, Level 1 loads and Level 2 loads are considered important loads. When restoring power, ensure that all primary loads are restored and that secondary loads are restored to the greatest extent possible.

8. A distribution network fault recovery system based on load-side islanding for implementing the distribution network fault recovery method according to any one of claims 1-7, characterized in that, include: The islanding module is used to obtain a connectivity graph based on the distribution network topology and adjust the edge weights. Starting from the primary load, it searches for initial islands. Using the initial islands as the search starting point, it searches for remaining loads to be assigned to islands using the edge weights, thus obtaining the distribution network islanding results and its objective function. The island fusion module is used to calculate the power supply redundancy of each island in the island division result, and to perform island fusion based on the power supply redundancy of each island to obtain the island fusion result. And based on the objective function of the island partitioning result, the objective function of island fusion result is obtained; The power supply restoration module for the distribution network load is used to obtain the maximum load restoration amount of the distribution network based on the islanding fusion results, its objective function, and constraints when the distribution network fails, so as to restore power supply to the load.

Images